part ii - research.vumc.nl 3.pdf · hd patient serum had a higher capacity to synchronize cells in...
TRANSCRIPT
Part IIDisturbed circadian rhythms
Chapter 3Disturbed peripheral circadian rhythms
in hemodialysis patients
Marije Russcher* Inês Chaves*Karolina Lech
Birgit C.P. KochJ. Elsbeth Nagtegaal
Kira F. DorsmanAnke ’t Jong
Manfred KayserH. (Martijn) J. R. van Faassen
Ido P. KemaGijsbertus T.J. van der Horst
Carlo A.J.M. Gaillard
* Both authors contributed equally to this study
submitted
3
Chapter 3
54
AbstractThe quality of life of hemodialysis (HD) patients is hampered by reduced nocturnal sleep quality and
excessive daytime sleepiness. In addition to the sleep/wake cycle, levels of circadian biomarkers (e.g.
melatonin) are disturbed in end-stage renal disease (ESRD). This suggests impaired circadian clock
performance in HD patients, but the underlying mechanism is unknown. In this observational study,
circadian rhythms of sleep, melatonin, cortisol and clock gene expression are compared between HD
patients (n=9) and healthy age-matched control subjects (n=9). In addition, the presence of circulating
factors that might affect circadian rhythmicity is tested. Reduced sleep quality and increased daytime
sleepiness were confirmed in HD patients. Reduced nocturnal melatonin concentrations and affected
circadian cortisol rhythms were found, as well as disrupted circadian expression of the clock gene
REV-ERBα. HD patient serum had a higher capacity to synchronize cells in vitro, suggesting an
accumulated level of clock resetting compounds in HD patients. These compounds were not cleared
by hemodialysis treatment or related to frequently used medications. In conclusion, the above
mentioned results strongly suggest a disturbance in circadian timekeeping in peripheral tissues of
HD patients. Accumulation of clock resetting compounds possibly contribute to this. Future studies
are needed for a better mechanistic understanding of the interaction between renal failure and
perturbation of the circadian clock.
Disturbed peripheral circadian rhythms in hemodialysis
55
3
Introduction
The quality of life of hemodialysis (HD) patients is hampered by reduced nocturnal sleep
quality and excessive daytime sleepiness.(1,2) Melatonin plays an important role in the
sleep wake rhythm.(3) It is an important biomarker of the body’s circadian clock. Melatonin
levels show a circadian rhythm with low levels during the day and high levels at night,
thereby providing a ‘nighttime’ signal to the body.(4–6) Levels of melatonin are remarkably
lowered in HD patients.(7,8) In addition to the sleep/wake rhythm and melatonin rhythm,
other diurnal rhythms, e.g. blood pressure (9) are disturbed in end-stage renal disease
(ESRD). Chronic circadian rhythm disruption has been associated with an increased risk to
develop morbidities such as obesity, diabetes, cardiovascular events and cancer (10–14),
but the causes of circadian rhythm disturbances in HD patients are unknown.
The human circadian system confers temporal organization upon behavior, physiology
and metabolism and enables the body to anticipate and adapt to daily recurring changes
in the environment.(15,16) It is composed of a central clock in the suprachiasmatic nuclei
(SCN) in the hypothalamus in the brain and peripheral clocks in virtually all other cells
and tissues.(17) Circadian rhythms are generated by an intracellular molecular oscillator,
composed of a set of clock genes and proteins, with an intrinsic period of approximately
24 hours. To keep synchronized with the earth’s day-night cycle, the SCN receives clock-
resetting environmental light information from the eye via the retino-hypothalamic tract.
(18) In turn, peripheral oscillators are synchronized by the central SCN pacemaker through
humoral and neuronal signals.(17,19)
To date, little is known about circadian clock performance in ESRD patients. One pre-
clinical study with nephrectomy-induced CKD in rats associated sleep disruption with
changes in the expression of clock genes (i.e. increased rPer1 and rPer2 mRNA levels)
in the hypothalamus (20), but the origin of the altered clock gene expression patterns
remained unsolved. Likewise, it is not known to what extent changes in serum metabolite
levels, inherent to the severely impaired kidney function in HD patients, impacts on
circadian clock performance.
The aim of the present study was to compare circadian clock performance in HD patients
with that of healthy, age and sex-matched control subjects. To this end, we analyzed sleep
quality and melatonin and cortisol rhythms, both important hormonal biomarkers of the
clock.(21) Moreover, we assessed circadian clock performance by measuring the expression
profiles of four circadian clock genes (Period (PER) 1, PER3, REV-ERBα and BMAL1) in
peripheral leukocytes. Finally, we explored whether alterations in serum metabolite
composition of HD patients circulating factor may influence peripheral circadian rhythms.
Chapter 3
56
Materials and Methods
Subjects
Participant inclusion and blood sampling was conducted at Meander Medical Centre,
Amersfoort, The Netherlands from July 2012 to October 2013. The study protocol was
approved by the institutional review board (Netherlands Trial Registry: NTR3631) and
written informed consent was obtained from all participants. The study was conducted
according to the Declaration of Helsinki. HD patients and community-based healthy
volunteers aged 18 to 85 years were eligible for inclusion. HD patients should be on
stable, chronic hemodialysis treatment for at least 3 months. Control person should have
MDRD-eGFR >60 ml/min/1.73m2. Participants were excluded in case of (i) blindness, (ii)
fever (>38ºC), (iii) C-reactive protein > 20 mg/L, (iv) use of hypnotics during the study
period, (v) exposure to anesthesia within 48 hours prior to start study, or (vi) serious co-
morbidity that prevented participation in this study according to the investigators.
Study design
Participants were allowed to maintain their individual sleep-wake rhythm and food intake
prior to and during the study (see figure 1 for study design). Participants entered the
hospital at 8:30 AM on day 1 and remained there until 9:30 AM on day 2. They spent the
overnight period in an individual room. Blood samples were obtained by venepuncture
at 9:00 (day 1), 13:00, 17:00, 21:00, 01:00, 05:00 and 9:00 (day 2) and collected in
regular serum collection tubes for serum analysis. In parallel, blood was collected into
PAXgeneTM RNA tubes (BD, Erembodegem, Belgium), allowing direct stabilization of
intracellular RNA for gene expression studies. In addition, blood samples were taken from
HD patients just before and just after HD treatment.
FIGURE 1. Study design HD=hemodialysis treatment, = blood sampling (for RNA isolation and serum collection, both HD patients and controls), = serum collection (for HD patients only)
Disturbed peripheral circadian rhythms in hemodialysis
57
3
Sleep analysis
Sleep parameters were investigated by means of actigraphy, using an Actiwatch 2 device
(Respironics Inc., Murrysville, Pennsylvania, USA). Actigraphy is an established sleep
monitoring method that records wrist movements and automatically discriminates rest-
activity patterns interpreted in terms of sleep and wake periods.(22) In addition, participants
were asked to record bedtimes and rise times on a registration form. Respironics Actiware
version 5.59 was used to score 1 minute epochs of actigraphic data as sleep or wake.
The following parameters were calculated according to standardized methods: ‘sleep
onset latency’ (SOL; time period between ’lights off’ and sleep onset), ‘sleep efficiency’
(SE; actual sleep time divided by time in bed), ‘actual sleep time’ (AST; total duration of
recorded sleep periods) and ‘wake after sleep onset’ (WASO; amount of time spent awake
after sleep has been initiated and before final awakening). Each episode of actigraphy
recordings was carried out during 4 consecutive days and nights, including HD and non-
HD days. Daytime sleepiness scores were measured by the Epworth Sleepiness Scale
(ESS). This questionnaire measures daytime sleepiness and has been used in HD patients
before.(2) Morningness-eveningness of the participants was determined by a chronotype
questionnaire, adapted from the Horne and Ostberg Owl and Lark questionnaire.(23)
Hormone assays
Cortisol and melatonin concentrations were measured with online-solid phase extraction
in combination with isotope dilution LC-MS/MS. In short, 200 μL of plasma was used for
the analysis and deuterated cortisol and melatonin were used as internal standard. Mean
intra- and inter-assay coefficients of variation were below 10%. Quantification limits for
cortisol and melatonin were 0.2 nM and 8.0 pM respectively.
Gene expression studies
RNA was isolated from leukocytes using the PAXgeneTM Blood miRNA kit (Qiagen,
Venlo, The Netherlands) according to the manufacturer’s instructions. cDNA synthesis
was performed using the RevertAid H Minus First Strand cDNA Synthesis Kit (Thermo
Scientific, Rochester, NY) and random hexamer primers, as described in the manufacturer’s
instructions. The qRT-PCR reaction was performed using the Lightcycler 480 SYBR Green
I Kit on a LightCycler 480 II platform (Roche Applied Bioscience, Indianapolis, IN). The
protocol was as follows: 10 minutes at 95 ºC, 45 cycles of 95 ºC, 60 ºC and 72 ºC (10
seconds each), followed by a melting curve step with continuous acquisition from 65 ºC
to 97 ºC. Relative quantification of gene expression was performed using the second
derivative maximum method (Roche Diagnostics), followed by the delta-delta-cycle-
Chapter 3
58
threshold (2-ΔΔCT) as described.(24) Clock gene expression of robust human clock genes
PER1, PER3, REV-ERBα, BMAL1, normalized to the endogenous housekeeping gene
β-actin (ACTB) were determined. Primers were acquired from Metabion (Martinsried,
Germany), tested for efficiency and specificity of amplification. Primer sequences used:
PER1: Fwd 5’-GGA CAC TCC TGC GAC CAG GTA CTG-3’, Rev 5’-GGC AGA GAG GCC
ACC ACG GAT-3’; PER3: Fwd 5’-GGT CGG GCA TAA GCC AAT G-3’, Rev 5’-GTG TTT
AAA TTC TTC CGA GGT CAA A-3’; ACTB: Fwd 5’-TGA CCC AGA TCA TGT TTG AG-3’,
Rev 5’-CGT ACA GGG ATA GCA CAG-3’; BMAL1: Fwd 5’-GCC TAC TAT CAG GCC AGG
CTC A-3’, Rev 5’-AGC CAT TGC TGC CTC ATC ATT AC-3’; REV-ERBα: Fwd 5’-GCG GCG
ATC GCA ACC TCT AG-3’, Rev 5’-GTA GGT GAT GAC GCC ACC TGT GTT-3’. All gene
expression levels at all time points of the participants were calculated relative to the gene
expression of the particular gene at the time point 13:00 of the first control person.
Cell culture and real time imaging of clock performance
HEPA1-6 cells, stably expressing a lentiviral Per2::luciferase clock reporter construct were
cultured in DMEM medium, supplemented with ultra-glutamine and 10% fetal calf serum
(Lonza, Basel, Switzerland). For in vitro testing of the clock synchronizing capacity of
patients and control serum, HEPA1-6 cells were seeded in 30 mm dishes in culture medium
supplemented with HEPES (25 mM final concentration; Lonza, Basel, Switzerland) and
luciferin (10 mM final concentration) and allowed to attach and stabilize for one day. Next,
dishes were placed in a LumiCycle bioluminescence imager to monitor clock performance
(see below). When the intracellular clocks in the cell population were desynchronized (72
hour after seeding), serum from HD patients or control subjects was added to the medium
and cultures were monitored for the reappearance of bioluminescence rhythms. As a
positive control, cells were synchronized with dexamethasone (100 nM final concentration).
Likewise, the clock synchronizing capacity of frequently used drugs was tested. To this end
pharmacological concentrations of acenocoumarol, acetaminophen, acetylsalicylic acid,
amlodipine, furosemide, insulin, metoprolol, nadroparin and oxazepam were added to
pooled human serum from the blood bank Erasmus MC. In all experiments, the amplitude
(peak/through) of the oscillation was used as a readout for synchronization potential.
Real time bioluminescence recordings of luciferase activity (60 seconds measurements
at 10 minute intervals) were performed using a LumiCycle 32-channel automated
luminometer (Actimetrics), placed in a dry incubator at 37 ºC. Prior to placing dishes in
the LumiCycle, the lid was replaced by a glass coverslip and sealed with Parafilm. Data
were analyzed using Actimetrics software.
Disturbed peripheral circadian rhythms in hemodialysis
59
3
Data analysis
Mean values and standard deviations of actigraphy, ESS and morningness-eveningness
scores, melatonin concentrations per time point, and in vitro amplitudes of bioluminescence
were calculated and/or plotted. Comparisons between control and patient results were
made with Student’s T test.
Individual mRNA expression profiles and cortisol curves were subjected to analysis by
the CircWave V1.4 programme (http://www.euclock.org/results/item/circ-wave.html). This
program determines the significance of circadian oscillation of an individual curve. Because
of its square wave pattern (25), this circadian analysis was not appropriate for the melatonin
curves.
Results
Nine HD-patients and 9 age- and sex-matched control subjects completed the study
protocol and were included in the analysis. The clinical characteristics of patients and
controls are summarized in table 1. As expected, hemoglobin levels were lower and
creatinin levels were higher in HD patients compared with controls. In line with the notion
that diabetes mellitus is frequently related to the development of ESRD (26), the HD
group contained more diabetic patients.
Sleep is affected in HD patients
We confirmed previous results showing that sleep is disturbed in HD patients.(27) HD-
patients suffered from a longer sleep onset latency (mean 25.1 ± 13.7 minutes versus 8.9
± 8.5 minutes, p=0.01), a lower sleep efficiency (mean 70.2 ± 8.1% versus 82.9 ± 10.9%,
p=0.02) and a higher number of awake minutes during the night after initial sleep onset
(mean 104.8 ± 27.9 minutes versus 54.6 ± 41.6 minutes, p=0.01), when compared with
healthy controls. Furthermore, HD patients reported elevated daytime sleepiness. Mean
Epworth Sleepiness Scale scores (± SD; a score >9 indicating elevated daytime sleepiness)
were 10.0 ± 4.8 and 3.9 ± 2.0 respectively (p<0.005). The morningness-eveningness
scores, as assessed by the chronotype questionnaire, did not differ significantly between
patients and controls.
Chapter 3
60
TABlE 1. Clinical characteristics
Hemodialysis patients Control persons p-value
Male / female 9 / 0 9 / 0
Age (years) 65.4 ± 12.1 59.4 ± 12.3 0.31
Body Mass Index (kg/m2) 28.3 ± 5.3 25.9 ± 3.1 0.27
Hemoglobin (mmol/L) 7.0 ± 0.5 9.4 ± 0.6 <0.001
Creatinin (μmol/L) 1021 ± 229 88.4 ± 6.9 <0.001
eGFR* (ml/min/1.73m2) hemodialysis 82.1 ± 10.1
Dialysis (hrs/week) 10.8 ± 2.0 n.a.
Smoking 3 1 0.58#
Diabetes mellitus 6 0 0.009#
ESS 10.0 ± 4.8 3.9 ± 2.0 <0.005
Chronotype questionnaire score 19.0 ± 3.1 21.7 ± 2.9 0.08
Sleep total (min.) 364 ± 92 381 ± 60 0.67
Sleep onset latency (min.) 25.1 ± 13.7 8.9 ± 8.5 0.01
Sleep efficiency (%) 70.2 ± 8.1 82.9 ± 10.9 0.02
WASO (min.) 104.8 ± 27.9 54.6 ± 41.6 0.01 Data are expressed as mean ± SD or number of patients where appropriate.* Calculated by the CKD-epi formula# Fisher’s exact test, http://in-silico.net/tools/statistics/fisher_exact_test, accessed on 29th Dec 2013eGFR = estimated Glomerular Filtration Rate, ESS = Epworth Sleepiness Scale, WASO = wake after sleep onset (in minutes)
Melatonin and cortisol oscillations are affected in HD patients
Being important outputs of the circadian clock, we have measured cortisol and melatonin
rhythms in the serum of HD patients and control subjects (figures 2A and B). As expected,
mean melatonin peak levels were higher in control subjects than in HD patients. At 1 AM,
the mean melatonin concentration of the control persons was significantly higher than
that of the HD-patients (91.9 ± 61.8 pmol/L versus 29.7 ± 41.2 pmol/L, p=0.025) and
at 5 AM, their mean melatonin concentration tended to be higher (111.9 ± 86.3 pmol/L
versus 39.6 ± 53.3 pmol/L, p=0.051). Likewise, cortisol rhythms differed between patients
and controls. Significant circadian oscillation of cortisol was found in 2 HD-patients and in
6 control subjects. At 5 AM, the mean cortisol concentration of the control persons was
significantly higher than that of the HD-patients (251 ± 83 nmol/L versus 169 ± 54 nmol/L,
p=0.026). These findings are in line with previously reported circadian melatonin and
cortisol measurements in ESRD patients, demonstrating reduced nocturnal melatonin
concentrations (7,8,28) and elevated cortisol nadir in the evening (29), as compared to
Disturbed peripheral circadian rhythms in hemodialysis
61
3
healthy subjects. Moreover, our data suggest that the cortisol and melatonin rhythms are
less robust in HD patients.
Circadian expression of REV-ERBα is disturbed in HD patients
Given the reduced amplitude of melatonin and cortisol rhythms in HD patients, we next
questioned whether circadian rhythmicity in peripheral clock gene expression is affected
in HD patients. To this end, we measured PER1, PER3, REV-ERBα and BMAL1 mRNA
rhythms in white blood cells by quantitative RT-PCR. Analysis of individual gene expression
profiles of the control subjects (supplementary figure S1) revealed that REV-ERBα is the
most reliable circadian marker. Significant REV-ERBα mRNA rhythms were observed in
6 out of 9 control subjects, whereas significant PER1, PER3 and BMAL1 mRNA rhythms
were detected in only 2, 3 and 2 controls, respectively (figure 2C). Strikingly, none of the
HD-patients showed a circadian REV-ERBα mRNA rhythm (p=0.009). Moreover, REV-ERBα
mRNA, but also PER3 and BMAL1 mRNA levels were consistently lower in HD patients
(average values shown in figure 2D-G). Taken together, these data suggest that circadian
gene expression is markedly affected in peripheral blood cells of HD patients.
Serum from HD patients has a higher capacity to synchronize cells in vitro
Peripheral clocks are daily synchronized by the central SCN clock though cyclic release
of humoral factors.(19) Given the severely impaired kidney function in CKD patients, it
is tempting to speculate that abnormal levels of serum metabolites (including humoral
factors involved in peripheral clock synchronization) underlie altered levels in the blood
of CKD patients. We therefore explored the hypothesis that the dampened amplitude
of cortisol, melatonin and peripheral clock gene expression patterns may (in part) be
caused by impaired clearance of clock synchronizing factors by the kidney resulting
in increased, saturating levels of clock synchronizing factors. Under this condition,
peripheral cells constitutively, rather than cyclically are triggered to reset their clocks,
causing desynchronization of these peripheral clocks. To test this hypothesis, we have
made use of the phenomenon that asynchronous circadian clocks of cultured peripheral
cells (e.g. fibroblasts, hepatocytes) can be temporarily synchronized by addition of serum
to the medium (known as serum shock). The higher the serum concentration, the higher
will be the amplitude of synchronized oscillation.
Chapter 3
62
Disturbed peripheral circadian rhythms in hemodialysis
63
3
FIGURE 2. Melatonin, cortisol and clock gene rhythmsPanel A: mean oscillation (±SD) of melatonin. The horizontal axis reflects the time of day in hours and the vertical axis reflects melatonin concentration in serum (pM). The detection limit of melatonin was 8 pM. Individual results below the detection limit, were set at 0 pM. Panel B: mean oscillation (±SD) of cortisol. The horizontal axis reflects the time of day in hours and the vertical axis reflects cortisol concentration in serum (nM). Panel C: Characterization of circadian rhythmicity. The vertical axis reflects the percentage of individuals in whom a significant circadian rhythm was calculated with CircWave. The horizontal axis reflects each circadian marker. The actual number of individuals is shown above each bar. Panels D to G: Mean (±SD) oscillations in expression of clock genes PER1, PER3, REV-ERBα and BMAL1. The horizontal axes reflect the time of day in hours. The vertical axes reflect clock gene mRNA expression (relative copy number/β-actin, all arbitrarily normalized to the relative expression at 13:00 of the first control person). The black lines represent patients, the dashed lines represent controls. n=9 for both control and patient groups. * p<0.05
We first tested the effect of increasing concentrations of pooled human control serum on
the amplitude of bioluminescence rhythms in Hepa1-6 cells, expressing a Per2::luciferase
clock reporter gene. As shown in supplementary figure S2, human serum can dose
dependently synchronize the circadian clock. Importantly, when using serum from HD
patients, the amplitude of oscillation was higher at all serum concentrations tested, with
the most significant difference observed at 6% serum (supplementary figure S2). These
data suggest that HD patient serum has a higher clock synchronizing potential.
We next compared the clock synchronizing capacity of serum taken from control subjects
and HD patients from our study groups at a fixed concentration of 6% serum. In line with
the previous experiment, the serum of HD patients has a higher capacity to synchronize
the circadian clocks of cultured HEPA1-6 cells than serum of healthy controls (figure 3
upper panels), as reflected by the larger amplitude of bioluminescence (2.61 ± 0.61 and
1.89 ± 0.32 for patient and control sera respectively, p=0.008).
To eliminate the possibility that the observed higher synchronizing capacity was caused by
the medication of HD patients, commercially available human pooled serum was spiked
with pharmacological concentrations of frequently prescribed drugs (acenocoumarol,
acetaminophen, acetylsalicylic acid, amlodipine, furosemide, insulin, metoprolol,
nadroparin and oxazepam) and tested for modulation of clock synchronizing capacity.
None of the drugs tested was capable of synchronizing HEPA1-6 cells at pharmacological
concentrations (supplementary figure S3).
To test whether hemodialysis influences the synchronizing capacity of the patients’
serum, e.g. by removing clock resetting factors from the bloodstream, patient serum
was collected just before and right after hemodialysis. No significant difference in syn-
chronization capacities was observed (figure 3 lower panels), reflected in mean amplitudes
of 2.50 ± 0.52 versus 2.63 ± 0.65 before and after hemodialysis treatment, respectively.
Hemodialysis does not reduce the clock synchronizing capacity of serum.
Chapter 3
64
FIGURE 3. Cell synchronizationUpper panels (A): in vitro synchronization capacity of HD patients’ serum (n=9) versus controls’ serum (n=9) on HEPA1-6 cells carrying a Per2::luciferase reporter construct. Upper left panel: graphical representation of the amplitude of oscillation. A higher amplitude indicates a higher degree of synchronization. Upper right panel: mean baseline corrected bioluminescence recordings of controls (dashed line) and patients (solid line). Lower panels (B): in vitro synchronization capacity of HD patients’ serum (n=9) on HEPA1-6 cells carrying a Per2::luciferase reporter construct before and after hemodialysis treatment. Lower left panel: graphical representation of the amplitude of oscillation. A higher amplitude indicates a higher degree of synchronization. Lower right panel: mean baseline corrected bioluminescence recordings of serum before (solid line) and after (dotted line) dialysis.** p<0.01
Disturbed peripheral circadian rhythms in hemodialysis
65
3
Discussion
The present study assessed peripheral diurnal oscillations in human HD patients at a
behavioral, hormonal and molecular level. Patients and control subjects were allowed
to follow their own daytime routine inside and around the hospital in order to stay close
to “real life” conditions and preserve the naturally occurring variation. The disturbance
of the sleep wake cycle in HD patients was confirmed by lower nighttime sleep quality
and higher daytime sleepiness in addition to a reduced amplitude of melatonin and
cortisol rhythmicity. In order to identify alterations in peripheral circadian oscillators, we
compared the expression levels of four clock genes in leukocytes of HD-patients with
those of healthy age-matched control subjects. We observed a marked difference in the
circadian expression profile of patients compared to controls. Overall, PER3, BMAL1 and
REV-ERBα mRNA levels were lower in HD patient leukocytes, suggesting that intracellular
peripheral clocks are dampened. Moreover, when analyzing the presence of circadian
oscillations, we observed a statistically significant difference between HD patients and
control subjects for REV-ERBα rhythms, which did not show significant oscillation in any
of the patients. The other genes tested appeared less informative, as significant rhythms
could only be detected in a limited number of control subjects. Nonetheless, even for
BMAL1 and PER1, we could not detect significant mRNA rhythms in HD patients, further
supporting the dampening of the circadian clock in HD patients. Taken together, even
in the absence of a constant routine protocol (e.g. standardized light and meal times),
our behavioral, hormonal and gene expression data point to impaired circadian clock
performance in HD patients.
Apart from a reduction in clock gene expression in individual cells, at the population level
of cells, dampening of circadian amplitude can also originate from a desynchronization
of individual cellular clocks. We hypothesized that circadian dysregulation in HD patients
may (in part) be caused by elevated levels of circadian clock resetting compounds, which
accumulate in the blood from HD patients due to impaired renal clearance. In this scenario,
the central clock still functions normally, in HD patients, but clock synchronizing humoral
signals from the SCN are not sufficiently cleared and accumulate in the bloodstream at
concentrations that exceed threshold levels and accordingly continuously, rather than
rhythmically, trigger peripheral clocks to reset. This ultimately causes peripheral oscillators
to desynchronize, resulting in a lower amplitude, or even absence of peripheral circadian
rhythms, as found for melatonin, cortisol and REV-ERBα expression. The concept of the
presence of certain serum factors being responsible for circadian changes has already
suggested before in clinical research on increasing age.(30)
Chapter 3
66
In support of the above hypothesis, using an in vitro clock synchronization assay with
Per2::luciferase Hepa1-6 cells, we have shown that HD patient serum has a higher clock
resetting potential than control serum. As commonly used HD related drugs did not have
any influence on cell synchronization, the synchronizing factor(s) is (are) of endogenous
origin. Hemodialysis treatment did not reduce the synchronizing potential of patient
serum. An explanation for this observation could be that hemodialysis does not remove
these compounds to substantially lower concentrations (i.e. below threshold values).
Which serum components are responsible for the effect observed remains to be
determined. Since melatonin and cortisol synthesis and secretion as well as sleep/activity
all result from circadian regulation within peripheral (i.e. non-SCN) tissues, their changes
in HD patients might all share the same origin.
At first sight, our data seem to contradict the preclinical study of Hsu and coworkers
(20), in which sleep disruption in nephrectomized rats was shown to be associated with
a significant upregulation of rPer1 and rPer2 expression in the hypothalamus. Yet, closer
examination of the data suggest that, despite higher expression levels, the amplitude
of rPer1 and rPer2 rhythms is reduced, which would be in line with our study (figure
2). On the other hand, differences may originate from the fact that Hsu and coworkers
investigated clock gene expression in brain tissue (i.e. hypothalamus), which apart from
humoral signaling, is also subject to neuronal signaling from the SCN.
Our results point to perturbation of circadian control in HD-patients, but it cannot be
ascertained that the observed differences are solely renal function related. Although
sleep deprivation itself influences clock gene expression (31), it is not likely that disturbed
sleep is fully responsible for our observations. Diabetes could as well be a confounding
factor, as most HD patients and none of the controls have diabetes. This reflects the
well-known clinical situation in which end-stage renal disease occurs as a consequence of
long-term poorly regulated glucose levels. However, both at gene expression level and
synchronizing potential the patients without diabetes do not cluster closer to the controls,
suggesting that it is not diabetes that is causing the rhythm alterations. This is further
validated by the following observations: i) serum with higher glucose concentrations does
not induce a higher amplitude of oscillation (supplementary figure S4) and ii) insulin does
not induce a significant circadian oscillation (supplementary figure S3).
In conclusion, we have identified disturbances in circadian rhythmicity of sleep, mela-
tonin, cortisol and clock gene expression in HD patients. Notably, clock gene expression
(including circadian amplitude) in peripheral cells was shown to be reduced in HD
patients. Furthermore, our data suggest that impaired peripheral clock performance
may originate from compromised humoral clock resetting from the SCN. Serum of
Disturbed peripheral circadian rhythms in hemodialysis
67
3
HD patients has a higher clock synchronizing capacity than serum of matched control
subjects. This strongly suggests a disturbance in circadian timekeeping in HD patients,
induced by reduced renal clearence. Further studies are needed in order to understand
the mechanism of the interaction between renal failure and perturbation of the circadian
clock. Importantly, efforts should be made to determine the nature of accumulated clock
resetting compounds. Understanding the origin of circadian rhythm disturbances in HD
patients is important in order to further optimize current treatment strategies and to
identify novel therapeutic targets.
Declaration of interest
The authors of this manuscript have no conflicts of interest to disclose.
Acknowledgements
The authors would like to thank Jeannette Schraa and Adry Diepenbroek for their clinical
assistance. This work was supported in part by a Leo Fretz foundation grant.
Chapter 3
68
References 1. Parker KP, Bliwise DL, Bailey JL, Rye DB. Daytime sleepiness in stable hemodialysis patients.
Am J Kidney Dis. 2003;41(2):394–402. 2. Parker KP, Kutner NG, Bliwise DL, Bailey JL, Rye DB. Nocturnal sleep, daytime sleepiness, and
quality of life in stable patients on hemodialysis. Health Qual Life Outcomes 2003;1:68. 3. Arendt J. Melatonin: characteristics, concerns, and prospects. J Biol Rhythms 2005;20(4):291–
303. 4. Claustrat B, Brun J, Chazot G. The basic physiology and pathophysiology of melatonin. Sleep
Med Rev. 2005;9(1):11–24. 5. Arendt J. Melatonin and human rhythms. Chronobiol Int. 2006;23(1-2):21–37. 6. Brzezinski A. Melatonin in humans. N Engl J Med. 1997;336(3):186–95. 7. Koch BCP, Nagtegaal JE, Hagen EC, van der Westerlaken MML, Boringa JBS, Kerkhof GA,
et al. The effects of melatonin on sleep-wake rhythm of daytime haemodialysis patients: a randomized, placebo-controlled, cross-over study (EMSCAP study). Br J Clin Pharmacol. 2009;67(1):68–75.
8. Russcher M, Koch BCP, Nagtegaal JE, van Ittersum FJ, Pasker-de Jong PCM, Hagen EC, et al. Long-term effects of melatonin on quality of life and sleep in haemodialysis patients (Melody study): a randomized controlled trial. Br J Clin Pharmacol. 2013;76(5):668–79.
9. Covic A, Goldsmith DJ. Ambulatory blood pressure monitoring in nephrology: focus on BP variability. J Nephrol. 1999;12(4):220–9.
10. Doi M, Takahashi Y, Komatsu R, Yamazaki F, Yamada H, Haraguchi S, et al. Salt-sensitive hypertension in circadian clock-deficient Cry-null mice involves dysregulated adrenal Hsd3b6. Nat Med. 2010;16(1):67–74.
11. Gumz ML, Stow LR, Lynch IJ, Greenlee MM, Rudin A, Cain BD, et al. The circadian clock protein Period 1 regulates expression of the renal epithelial sodium channel in mice. J Clin Invest. 2009;119(8):2423–34.
12. Stow LR, Gumz ML. The circadian clock in the kidney J Am Soc Nephrol. 2011;22:598–604. 13. Philips ML. Of owls, larks and alarm clocks. Nature 2009;458:142–4. 14. Guo Y, Liu Y, Huang X, Rong Y, He M, Wang Y, et al. The effects of shift work on sleeping quality,
hypertension and diabetes in retired workers. PLoS One 2013;8(8):e71107. 15. Ralph MR, Foster RG, Davis FC, Menaker M. Transplanted suprachiasmatic nucleus determines
circadian period. Science 1990;247(4945):975–8. 16. Firsov D, Bonny O. Circadian regulation of renal function. Kidney Int. 2010;78(7):640–5. 17. Fukuhara C, Tosini G. Peripheral circadian oscillators and their rhythmic regulation. Front Biosci.
2003;8:642–51. 18. Sadun AA, Schaechter JD, Smith LE. A retinohypothalamic pathway in man: light mediation of
circadian rhythms. Brain Res. 1984;302(2):371–7. 19. McNamara P, Seo SB, Rudic RD, Sehgal A, Chakravarti D, FitzGerald GA. Regulation of CLOCK
and MOP4 by nuclear hormone receptors in the vasculature: a humoral mechanism to reset a peripheral clock. Cell 2001;105(7):877–89.
20. Hsu C, Chang F, Ng H, Kuo C, Lee Y, Lu C, et al. Disrupted circadian rhythm in rats with nephrectomy-induced chronic kidney disease. Life Sci. 2012;91:127–31.
21. Klerman EB, Gershengorn HB, Duffy JF, Kronauer RE. Comparisons of the Variability of Three Markers of the Human Circadian Pacemaker. J Biol Rhythms 2002;17(2):181–93.
22. Morgenthaler T, Alessi C, Friedman L, Owens J, Kapur V, Boehlecke B, et al. Practice parameters for the use of actigraphy in the assessment of sleep and sleep disorders: an update for 2007. Sleep 2007;30(4):519–29.
Disturbed peripheral circadian rhythms in hemodialysis
69
3
23. Kerkhof GA. Een Nederlandse vragenlijst voor de selectie van ochtend- en avondmensen. Ned Tijdschr voor Psychol. 1984;39:281–94.
24. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 2001;25(4):402–8.
25. Lewy AJ, Sack RL. The dim light melatonin onset as marker for circadian phase position. Chronobiol Int. 1989;6(1):93–102.
26. Broumand B. Diabetes: changing the fate of diabetics in the dialysis unit. Blood Purif. 2007;25:39–47.
27. Parker KP. Sleep disturbances in dialysis patients. Sleep Med Rev. 2003;7(2):131–43. 28. Karasek M, Szuflet A, Chrzanowski W, Zylinska K, Swietoslawski J. Decreased melatonin
nocturnal concentrations in hemodialyzed patients. Neuroendocr Lett. 2005;26(6):653–6. 29. Raff H, Trivedi H. Circadian rhythm of salivary cortisol, plasma cortisol, and plasma ACTH in
end-stage renal disease. Endocr Connect. 2013;2:23–31. 30. Pagani L, Schmitt K, Meier F, Izakovic J, Roemer K, Viola A, et al. Serum factors in older
individuals change cellular clock properties. PNAS. 2011;108(17):7218–23. 31. Ackermann K, Plomp R, Lao O, Middleton B, Revell VL, Skene DJ, et al. Effect of sleep
deprivation on rhythms of clock gene expression and melatonin in humans. Chronobiol Int. 2013;30(7):901–9.
Chapter 3
70
Supplementary data
FIGURE S1. Individual oscillations in expression of clock genes PER1, PER3, REV-ERBα and BMAl1 The horizontal axes reflect the time of day in hours. The vertical axes reflect clock gene mRNA expression (relative copy number/β-actin, all arbitrarily normalized to the relative expression at 13:00 of the first control person). The black lines represent patients, the dashed lines represent controls, n=9 for both groups.
Disturbed peripheral circadian rhythms in hemodialysis
71
3
FIGURE S2. Determination of optimal serum concentration to detect differences in in vitro synchronization capacity between hemodialysis patient serum (n=1, black dots) and pooled human control serum (grey diamonds)A larger amplitude indicates a higher degree of synchronization. * p < 0.05, ** p < 0.01
Chapter 3
72
FIGURE S3. Mean bioluminescence recordings of HEPA1-6 cells carrying a Per2::luciferase reporter construct Cells are synchronized with pooled human serum (panel A, mock treatment), serum containing dexamethasone 1 μM (panel B, positive control) and serum spiked with pharmaceutical drug concentrations (Panels C to J) (n=2). Panels K and L: mock treatment and insulin treatment respectively.
Disturbed peripheral circadian rhythms in hemodialysis
73
3
FIGURE S4. Mean amplitude results of bioluminescence recordings of HEPA1-6 cells carrying a Per2::luciferase reporter construct Cells are synchronized with hemodialysis patient serum spiked with increasing glucose concentration. A higher amplitude indicates a higher degree of synchronization. There was no significant correlation between glucose concentration and amplitude.